Results 8 comments of F+arming

> [7fa1387](https://github.com/Rayhane-mamah/Tacotron-2/commit/7fa13872398ae5a52c4138752416950258f5ae04) thank you very much, I'll try by this way. : )

多谢多谢,果然里面两句话就搞定了。 for find_re, replace in BOPOMOFO_REPLACE: pinyin = find_re.sub(replace, pinyin) pinyin = ''.join(BOPOMOFO_TABLE.get(x, x) for x in pinyin)

Hi, Zengqi-Qin and your team, thanks for your project. The model cards on HF (https://huggingface.co/myshell-ai/MeloTTS-Chinese/tree/main) is about 200 MB, but the model saved during my training is about 610 MB....

Dear contributor, I have tried many times (using different tools to download that zip or downloading it in python/jupyter), while the results always be "Time out". I can not open...

`sf.write(file_name, data, samplerate)` Maybe the data is not right, **numpy.ndarray** with multiple dimensions. such as '[[0,1,......]]' `sf.write(file_name, data[0], samplerate)`

"fp16_run": false时,训练的精度还是FP16精度对齐吗?还是FP32?

好久之前的问题了,有两点需要注意:1.整理音频数据,较长的音频数据需要较大显存。2. txt文件中的文字不要有生僻字或者生僻符号。希望对大家有帮助。 @Hjlvx @KexingWestRoad